The Heisenberg Equity Principle

In my more grandiose moments I imagine that I’m Charly Ellis, principal of Greenwich Associates and advisor to some of the nation’s wealthiest individuals, funds, and endowments. He’s also the author of The Loser’s Game, a true investment classic published in Financial Analysts Journal in 1975. (Long story short: Portfolio management at all levels resembles amateur tennis, in which the winner typically is not the player who makes the most brilliant shots, but rather the one making the fewest mistakes. In short—buy, hold, and do not trade unless absolutely necessary.)

I was thus surprised to catch him repeating in his 1998 book, Winning the Loser’s Game (McGraw-Hill), an old bromide which should have been long ago consigned to an anthology of investment mythology:

" . . . the typical pension fund was, in the 1970s, 1980s, and 1990s, only 50 percent in equities. . . . . These funds paid an opportunity cost in returns foregone. As it turned out, the cost of not being fully invested in stocks in the 1980s and 1990s was very large . . . . ."

In other words, had the nation’s largest investment pools invested more heavily in stocks, they would have reaped the high equity returns of the past two decades.

Such observations are not new. In an infamous interview in Ladies Home Journal in 1929 entitled "Everyone Ought to be Rich," John Raskob proposed that if everyone simply invested $15 in the stock market each month, before very long we would all become rich. Unfortunately, his computations imputed an annualized return of 24%. These days we’re much more sophisticated. Proponents of social security privatization are careful to assume the historical 8% real return for accounts invested in equities. Happily, even at this lower return, everybody still gets rich.

The mistake made by Ellis, Raskob, and privatization enthusiasts is subtle but important. Imagine for a moment that in 1980 all the nation’s pension funds and other long-lived investment pools shifted to 100% equity. What would obviously have occurred is a one-time pop in equity prices, followed by much lower subsequent returns, since half of equity returns are from dividends. (I’m indebted to Jason Zweig of Money for pointing out to me that this is precisely what happened in Chile, which experienced spectacular stock performance due to retirement privatization in the early 1980s, with lousy returns since.)

The problem can be thought of in another way. Since 1926 the real return of corporate equity has been about 8%, while that of corporate bonds has been about 2%. Which means that the aggregate return on corporate capital was somewhere in the vicinity of about 5%. Imagine for a moment that in 1926 Congress had made bonds illegal, forcing all investment/corporate capitalization into stocks. Equity prices would have initially been bid up, lowering subsequent returns as dividends fell. More importantly, companies would have been forced to issue huge amounts of equity for capitalization, diluting shareholder cash flow streams, lowering returns yet further. Let’s call this scenario "stockworld."

Now imagine the opposite had occurred—that in 1926 Congress had outlawed equity, leaving bonds as the only source of investment/capitalization—"bondworld." Here things get even more interesting. You can’t "dilute" interest payments. (Well, actually, you can. But that’s called default.) So the bond market becomes flooded with supply, driving down prices and forcing up yields. Even worse, the companies become very highly leveraged, lowering their credit quality, forcing up yields even more. Hello 5% real bond yields.

In fact, bondworld may be a preferable state of affairs to stockworld, since the former is very highly leveraged and the latter completely deleveraged. In bondworld there is zero margin for error in corporate capital allocation, and since all of the company’s profits flow to its private owners, the usual agency problems arising from differences between shareholder and manager goals cease to exist—there are no shareholders. In stockworld additional capital is more easily available via additional stock issuance, with much less incentive for management to use it efficiently.

In short, the aggregate national investment return will be approximately the same no matter what the overall stock/bond mix of the capital markets. To the extent that debt tends to decrease agency conflicts, a small nod may go to an increase in the overall debt/equity ratio. If everybody issues/invests in stocks, then stock returns must fall to the aggregate return rate. Which may actually already have happened. If all of the nation’s pension funds and newly-privatized social security accounts shifted to stocks, they most decidedly would not obtain the historical 7%-8% real return.

This situation is analogous to that of a physicist who attempts to measure the precise position of an electron. The laws of physics dictate that it cannot be done, since the measurement itself will displace the electron's original position—once we’ve ascertained that stocks have returns vastly superior to bonds, their prices will be rapidly bid up, and the previous expected equity risk premium reduced or eliminated. On the other hand, for opposite reasons, it seems likely that expected bond returns are much higher than actualized historical returns; perhaps even greater than that of stocks.

But one thing is clear: it’s not kosher to play the game of shoulda, woulda, coulda with a large slab of the nation’s investment pool. Had the majority of corporate pension officers known that stocks would beat bonds by 6% per year over the next several decades, it could not have happened. And now that tout le monde knows it, it won’t.

Copyright © 2000, William J. Bernstein. All rights reserved.